Research map / Bandits and online learning / Multi-armed bandits
A Change-Detection Based Framework for Piecewise-Stationary Multi-Armed Bandit Problem
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use PHT-UCB instead of the strongest reported baseline | Switching environment bandit policy · Switching environment | regret exponent b: 0.83 -> 0.69 | Helped |
| Use CUSUM-UCB instead of the strongest reported baseline | Yahoo! bandit policy with K=5 · Yahoo! dataset (K=5) | regret exponent b: 0.81 -> 0.69 | Helped |
| Use CUSUM-UCB instead of the strongest reported baseline | Yahoo! bandit policy with K=100 · Yahoo! dataset (K=100) | regret exponent b: 0.88 -> 0.85 | Helped |
| Replace CUSUM change detection with PHT | CD-UCB in the switching environment · Switching environment | regret exponent b: 0.72 -> 0.69 | Helped |
| Replace CUSUM change detection with PHT | CD-UCB on Yahoo! with K=5 · Yahoo! dataset (K=5) | regret exponent b: 0.69 -> 0.79 | Hurt |
Closest papers by meaning
- A Near-Optimal Change-Detection Based Algorithm for Piecewise-Stationary Combinatorial Semi-Bandits (AAAI 2020 · similarity 0.91)
- Constrained Feedback Learning for Non-Stationary Multi-Armed Bandits (NeurIPS 2025 · similarity 0.87)
- Adaptive Regret for Bandits Made Possible: Two Queries Suffice (ICLR 2024 · similarity 0.86)
- Fixed-Confidence Multiple Change Point Identification under Bandit Feedback (ICML 2025 · similarity 0.86)
- Non-Stationary Lipschitz Bandits (NeurIPS 2025 · similarity 0.86)
- Tracking Most Significant Shifts in Infinite-Armed Bandits (ICML 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.